Thierry Dudok de Wit
homepage Sent: Monday, August 22, 2011 11:17 AM
Here is a selection of topics I suggested today to José, leaving to him to decide what would fit best within the session. Some of it can easily go on a poster.
1) multiscale analysis (also involving an analysis by SVD) of the sunspot and group sunspot numbers, together with other solar proxies, in order to determine which of the former two a more appropriate for reconstructing tracers of solar activity. The focus is rather on the comparison between sunspot number and group sunspot number.
2) the use of the SVD for filling in a self-consistent way gaps in sunspot records from different observers. Although this topic is more data processing-oriented than the previous one, I believe it is relevant for learning how to stitch together different records as a prerequisite to building a reference sunspot number. A poster presentation would probably be more suitable.
3) blind source separation analysis of the butterfly diagram: the idea is to decompose the spatio-temporal variability of the butterfly diagram into different contributions (hence the wording 'blind source separation'). These contributions are well localised in latitude but are not fully hemispheric symmetric. Interestingly, their time-evolutions show marked differences from one solar cycle to the other, which highlights the importance of the latitudinal distribution of sunspots in the evolution of a solar cycle. From this you can then build a phase space representation of the solar cycle that allows to compare the dynamics of each cycle and provides much deeper insight than phase space representations based on the sunspot number only.